Preoperative Neutrophil-to-Lymphocyte Ratio as a Prognostic Predictor after Curative Resection for Hepatocellular Carcinoma
This study was designed to evaluate the impact of an elevated preoperative neutrophil-to-lymphocyte ratio (NLR) on outcome after curative resection for hepatocellular carcinoma (HCC).
Patients undergoing resection for HCC from January 1994 to May 2007 were identified from the hepatobiliary database. Demographics, laboratory analyses, and histopathology data were analyzed.
A total of 96 patients were identified with a median age at diagnosis of 65 (range, 15–85) years. The 1-, 3-, and 5-year overall survival rates were 80%, 58%, and 52%, respectively. Although the presence of microvascular invasion, NLR ≥5, and R1 resection margin were adverse predictors of overall survival, there were no independent predictors identified on multivariate analysis. The 1-, 3-, and 5-year disease-free survival rates were 74%, 63%, and 57%, respectively. Preoperative tumor biopsy, NLR ≥ 5, multiple liver tumors, microvascular invasion, and R1 resection margin were all predictors of poorer disease-free survival. Multivariate analysis showed that a NLR ≥ 5 and R1 resection margin were independent predictors of poorer disease-free survival. The median disease-free survival of those with a NLR ≥ 5 was 8 months compared with 18 months for those with a NLR < 5.
Preoperative NLR ≥ 5 was an adverse predictor of disease-free and overall survival.
Hepatocellular carcinoma (HCC) is one of the most common malignancies worldwide and its prevalence is increasing because of the increase in chronic hepatitis infection . Hepatic resection can potentially provide 5-year survival rates up to 70% in selected patients . The recently published American Association for the Study of Liver Diseases guidelines  have suggested that liver resection should be the treatment of choice for solitary lesions that arise in noncirrhotic livers and in patients with Child-Pugh grade A cirrhosis who have well-preserved liver function. Besides recognized adverse prognostic factors, such as large tumor size [4, 5], multifocal disease [6, 7], and vascular invasion [6, 8], there is increasing evidence that correlates the presence of systemic inflammation with poorer cancer-specific survival in certain tumors [9, 10]. Current understanding suggest that the host’s inflammatory response to tumor and/or the systemic effects exerted by the tumor leads to up-regulation of the inflammatory process, predisposes the tumor to proliferate and metastasize through the inhibition of apoptosis, promotion of angiogenesis, and damage of DNA [9, 11, 12].
The presence of a systemic inflammatory response can be determined by both the expression of C-reactive protein (CRP)  and an elevation in neutrophil-to-lymphocyte ratio (NLR) . Hashimoto et al.  showed that an elevated preoperative CRP was associated with early recurrence and poorer survival after resection for HCC. Similar findings have been observed in cases of colorectal cancer . Although an elevated CRP has been widely used in this respect, not all centers routinely measure preoperative CRP. An elevation in NLR, another marker of inflammation, has been shown to be associated with poorer prognosis in patients with cardiovascular and peripheralvascular disease [14, 17], colorectal carcinoma , and more recently, colorectal liver metastases . These studies have focused on the preinterventional or surgical profile of NLR as it reflects the inflammatory status of the patient before treatment. Although the white cell and differential counts may vary at different periods, a preoperatively high ratio of 5:1 is most likely to reflect aggressive disease and hence, poorer outcome.
This study was designed to evaluate the impact of systemic inflammation, represented by an elevated preoperative NLR, on outcome after potentially curative hepatic resection for HCC.
Materials and methods
Patients with HCC undergoing hepatic resection at the Hepatobiliary and Transplantation Unit, St. James’s University Hospital (SJUH), Leeds, United Kingdom, during the 13-year period from January 1994 to May 2007 were identified from a prospectively maintained hepatobiliary database.
Demographic, clinical, and laboratory data, including patient age, sex, white cell and differential counts, tumor markers [alpha (α)-fetoprotein], and presenting features were recorded. All white cell and differential counts were taken on the day before surgery, and none of the patients had clinical symptoms or signs of sepsis. The NLR was calculated from the differential count by dividing the absolute neutrophil count by the absolute lymphocyte count. NLR ≥ 5 was considered raised in accordance with published literature . Preoperative radiological assessments included a thoracic, abdomen and pelvis computed tomography (CT), isotope bone scan, and magnetic resonance imaging (MRI) of the liver.
Hepatic parenchymal transection was performed by using the Cavi-Pulse Ultrasonic Surgical Aspirator (CUSA, Model 200T, Valley Lab., Boulder, CO). The number of hepatic (Couinaud’s) segments resected was determined by the procedure performed as stated in the Brisbane nomenclature . In certain cases, an intermittent Pringle maneuver was used, consisting of repeated cycles of 15 minutes of ischemia followed by 5 minutes of reperfusion. Intraoperative ultrasound was performed to confirm the findings of preoperative imaging and to assist in surgical planning. Resection and reconstruction of the portal vein and inferior vena cava also was performed when the tumor was judged to have invaded the vasculature. A major resection was defined as resection of three or more segments. In cases in which multiple resections were performed at a single setting, the most extensive resection was considered the main procedure.
The postoperative data collated included complications, clinical outcome (morbidity and mortality), and histopathological information regarding the resected specimens. Patients who developed evidence of encephalopathy or became jaundiced, with coagulopathy after surgery, were considered as having developed significant transient liver failure. Perioperative mortality included all deaths within 30 days of surgery. Transfusion of blood products (packed red cells or whole blood) during surgery or in-hospital stay after surgery also was recorded.
Patients were followed up at specialist hepatobiliary clinics. An intensive policy of postoperative surveillance is practiced within this unit, and no patients were lost to follow-up. Patients have three monthly thoracic and abdominal CT performed during the first postoperative year, followed by six monthly during the second year. In the following three years (years 3–5), an annual CT scan was performed, and finally at years 7 to 10 of follow-up. In addition, liver MRI was used to define suspicious lesions demonstrated on CT and/or raised α-fetoprotein. Overall and disease-free survival data were recorded.
Categorical data were presented as frequency and were analyzed by using the Pearson’s chi-squared test. The Kaplan-Meier method was used to assess the actuarial survival and disease recurrence rate. Univariate analysis was performed to assess for a significant difference in clinicopathological characteristics that influenced overall and disease-free survival after curative resection. A multivariate analysis was performed by Cox regression (step-wise forward model) for variables significant on univariate analysis. All statistical analyses were performed by using the SPSS® for Windows™ version 15.0 (SPSS Inc., Chicago, IL), and statistical significance was taken at the 5% level.
During the study period, 96 patients underwent potentially curative resection for HCC. The median age at diagnosis was 65 (range, 15–85) years, and there were 72 men and 24 women. The most common mode of presentation was abdominal pain in 25 patients followed by malaise and weight loss in 10 patients. Preoperative α-fetoprotein was raised in 27 patients. The NLR was elevated at ≥5 in 26 patients before surgery.
Operative data of patients in this study
Operative data (n = 96)
Left hemihepatectomy (resection of segments 2, 3, 4 ± 1)
Right hemihepatectomy (resection of segments 5, 6, 7, 8 ± 1)
Left trisectionectomy (resection of segments 2, 3, 4, 5, 8 ± 1)
Right trisectionectomy (resection of segments 4, 5, 6, 7, 8 ± 1)
Bisegmentectomy (resection of 2 segments)
Nonanatomical liver resection
The overall morbidity was 41% and there were two in-hospital deaths. The most common complication was intra-abdominal sepsis (n = 9), followed by wound infection in five patients. Five patients developed liver insufficiency postoperatively and transient renal failure was reported in four patients.
The majority of tumors were 5 cm or larger in maximum diameter (n = 73). The tumor was multiple in 33 cases. There were 24 cases with microscopic evidence of tumor at the resection margin. Microvascular infiltration was detected in 50 cases. Fourteen patients had underlying cirrhosis.
Overall and disease-free survival
At present, 60 patients are alive at a median follow-up of 30 (range, 6–152) months. Recurrent disease was reported in 29 patients during the follow-up period.
Univariate analysis of overall and disease-free survival
Age ≥ 65 years
Biopsy of tumor
Viral hepatitis (B or C)
Major hepatic resection
Multiple (>1) tumors
Large tumor size (>50 mm)
Macroscopic vascular invasion
Microscopic vascular invasion
Normal liver parenchyma
Presence of cirrhosis
Positive resection margin
Multivariate analysis of variables influencing disease-free survival
Clinical variables for disease-free survival
Biopsy of tumor
NLR ≥ 5
Multiple (>1) tumors
Macroscopic vascular invasion
R1 resection margin
Univariate analysis of clinical and histopathology data comparing a neutrophil-to-lymphocyte ratio of <5 to ≥5
NLR < 5 (n = 70)
NLR ≥ 5 (n = 26)
Viral hepatitis (B or C)
Normal liver parenchyma
Presence of cirrhosis
Multiple (>1) tumors and/or large tumor size (>50 mm)
Microscopic and/or macroscopic vascular invasion
The presence of an association between systemic inflammation and poorer prognosis has been established for a number of tumors, especially in cases of colorectal carcinoma [10, 13, 16, 21]. Besides applying raised CRP as a marker of inflammation known to be of prognostic value in patients with colorectal carcinoma [10, 13], the presence of an elevated preoperative NLR also has been validated as a marker of inflammation and shown to be of prognostic significance in colorectal carcinoma . In the present study, NLR was selected as a marker for inflammation to be assessed, because patients treated within the institution did not routinely have CRP measured preoperatively. The current series recorded an elevation in NLR in 27% of patients undergoing liver resection for HCC. This prevalence is similar to that of patients with colorectal liver metastases treated within the institution with regards to inflammatory markers, including CRP  and NLR .
In the present series, poor prognostic indicators influencing overall survival included: preoperative NLR ≥ 5; microvascular invasion; and R1 resection margin. These patients had poorer overall survival on univariate analysis, although none of these factors were identified as independent predictors on multivariate analysis. With respect to disease-free survival, both preoperative NLR ≥ 5 and R1 resection margin were independent adverse predictors on multivariate analysis. This is the first study to implicate the relationship of an elevated preoperative NLR and a poorer prognosis for patients undergoing potentially curative liver resection for HCC. This association was present for both disease-free and overall survival. Because of the retrospective nature of this study and limited by the small sample size, further larger, prospective studies are required to validate this finding. In addition, there is a potential for the NLR profile to be used as a marker of disease response to treatment and recurrence. Studies assessing the correlation of serial levels of NLR before and after resection and at follow-up with recurrent disease and response to treatment are needed.
In addition, the current multivariate analysis also identified the status of the resection margin as a predictor of disease-free outcome. The institution has recently published its experience after resection of large HCC in 85 patients and showed adverse prognostic factors that influenced disease-free survival included R1 resection margin, multiple tumors, vascular invasion and preoperative tumor biopsy . The addition of an inflammatory response, represented by NLR, in a slightly larger series of 96 patients with updated follow-up, has decreased the influence of all the other variables except resection margin as independent adverse predictors. Other authors have identified vascular involvement [4, 8, 23, 24], larger tumor size [4, 25], number of tumors [26, 27], resection margin status , Child Pugh status , HCC type [7, 29], underlying cirrhosis , extent of resection [8, 31], and blood loss or transfusion [8, 32] as prognostic predictors for patients after liver resection for HCC. The small sample size of the present study may be the limiting factor that other clinicopathological factors did not have a significant impact on outcome after resection. Although an association has been identified between an elevated preoperative NLR and poorer outcome, patients who express such a response should still be considered for curative resection due to the limited therapeutic alternatives to resection that provide good long-term survival.
The present series also noted an association between the presence of an elevated NLR ≥ 5 and the presence of underlying cirrhotic liver parenchyma. However, other histopathological features were not significantly different between the two NLR groups. This could be explained by the fact that a high ratio of 5:1 was chosen as a cutoff because this value has already been validated in patients with colorectal cancer. Furthermore, the present finding requires further validation in a larger series of patients.
The association between elevated NLR and poor prognosis is complex and remains to be elucidated. One possible explanation is the host’s immune response to tumor is lymphocyte dependent. Several studies on patients with colorectal carcinoma and its corresponding metastases have demonstrated that patients with weaker lymphocytic infiltration at tumor margins have a worse prognosis [33, 34]. Hence, patients with an elevated NLR have a relative lymphocytopenia, and this may result in exhibiting a poorer lymphocyte mediated immune response to malignancy, thereby worsening their prognosis and increasing the risk of tumor recurrence. Another possible explanation is that a raised neutrophil count may aid in the development and progression of the neoplasm by providing an adequate environment for growth and proliferation. Circulating neutrophils are known to contain and secrete the majority of circulating vascular endothelial growth factor (VEGF), a proangiogenic factor that is thought to be involved in tumor development . This angiogenic activity has been associated with poor prognosis, as demonstrated in patients with gastric carcinoma [36, 37]. Recently, Park et al. showed that a strong VEGF-C (member of the VEGF family) expression on surgical specimens of intrahepatic cholangiocarcinoma significantly correlated with lymph node metastasis and was an independent predictor of a poorer prognosis . Because both NLR and CRP are markers of systemic inflammation, there is likely to be an association between NLR and CRP as similar effects are seen when raised. Xavier et al.  observed an increase in serum levels of VEGF in the presence of raised CRP concentrations. In addition, Canna et al.  observed an inverse correlation between CRP levels and tumor lymphocytic infiltration, with a raised CRP concentration indicative of a weak infiltration of lymphocytes at the periphery of the tumor. Patients with a raised NLR have a relative lymphocytopenia, and this may result in exhibiting a weaker lymphocyte-mediated immune response to the tumor, thereby worsening their prognosis.
The preoperative inflammatory status of patients with HCC, represented by NLR, could be potentially a valuable prognostic predictor in directing both preoperative and postoperative therapies to these patients to improve their survival outcome. Currently, there are no specific therapeutic modalities available for patients who express tumor-related inflammatory responses, although the antiangiogenic activities of cyclooxygenase-2 inhibitors [41, 42] and vaccines that promote lymphocyte response to tumor  are being evaluated. Further understanding in this area is required, which may lead to the development of preoperative therapeutic targets that influence the expression of tumor-related inflammatory responses, and this may improve survival outcomes after resection for HCC.
The presence of an elevated preoperative NLR has been identified as an adverse predictor of outcome in patients undergoing potentially curative resection for HCC. Patients with high preoperative NLR should be considered as candidates for additional therapies after resection.
- 6.Grazi GL, Cescon M, Ravaioli M, Ercolani G, Gardini A, Del Gaudio M, Vetrone G, Cavallari A (2003) Liver resection for hepatocellular carcinoma in cirrhotics and noncirrhotics. Evaluation of clinicopathologic features and comparison of risk factors for long-term survival and tumour recurrence in a single centre. Aliment Pharmacol Ther 17(Suppl 2):119–129PubMedCrossRefGoogle Scholar
- 40.Canna K, McArdle PA, McMillan DC, McNicol AM, Smith GW, McKee RF, McArdle CS (2005) The relationship between tumour T-lymphocyte infiltration, the systemic inflammatory response and survival in patients undergoing curative resection for colorectal cancer. Br J Cancer 92:651–654PubMedCrossRefGoogle Scholar
- 43.Lesterhuis WJ, de Vries IJ, Schuurhuis DH, Boullart AC, Jacobs JF, de Boer AJ, Scharenborg NM, Brouwer HM, van de Rakt MW, Figdor CG, Ruers TJ, Adema GJ, Punt CJ (2006) Vaccination of colorectal cancer patients with CEA-loaded dendritic cells: antigen-specific T cell responses in DTH skin tests. Ann Oncol 17:974–980PubMedCrossRefGoogle Scholar